连接操作是影响分布式查询性能的关键因素,数据存储是影响连接操作的重要因素。为了提高分布式系统的查询性能,通过研究数据之间的关系,提出一个关联数据分布树。利用该关联数据分布树来构造一系列的关联元组集合,然后按照各个站点的负载能力,把这些关联数据集合分配给相关站点。实验结果表明,当多个关系频繁的进行连接操作时,关联数据分布树能有效地提高整个分布式系统的查询性能。
Join operation is key factor that affects query performance. Data storage is the important factor affected join operation. To improve query performance of distributed system, a conjunction data distribution tree is proposed by studying relation between data. A series of conjunction tuple sets are constructed by the conjunction data distribution tree, then according to every site' s load capacity, these conjunction data sets are distributed to relative site. Experimental result shows that when multiple relations process join operation frequently, the conjunction data distribution tree can improve effectively query performance of total distributed system.